See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/312530110

DESIGNING AND TESTING OF TURBO-EXPANDER

Thesis · January 2017

DOI: 10.13140/RG.2.2.15846.01605

CITATIONS READS

0 3,962

1 author:

Hamad Khaled Alkhareef

University College London
8 PUBLICATIONS   1 CITATION   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Open for new assignment in reasearch View project

All content following this page was uploaded by Hamad Khaled Alkhareef on 19 January 2017.

The user has requested enhancement of the downloaded file.

 DESIGNING AND TESTING OF
TURBO EXPANDER
H1043 INDIVIDUAL PROJECT

University of Sussex
School of Engineering and Informatics
Department: Engineering and Design

BEng Mechanical Engineering

Supervisor: Dr. Vasudevan Kanjirakkad

Word count 8311 Candidate: 122511

 Summary

This project is about making and testing of an engineering application. A knowledge of

turbo-machinery and thermodynamics have involved to achieve requirements. The project is
based on designing and testing of a small turbo-expander for performance. The turbo-
expander is essentially consist of a centrifugal compressor and a radial inflow turbine
connected on a single rotating shaft. A mass flow rate of 0.048 kg/s is sucked into the
compressor and transfers directly to the radial turbine, the flow path is driven by a vacuum
cleaner, which is connected to the turbine exhaust. The ratio of the flow power at compressor
exit to the maximum suction power from the vacuum cleaner is 56% with an overall
efficiency of 35.9%. The maximum suction power from the vacuum cleaner is directly
proportional to the rotational speed.
This process produces energy, however, the similar flow path can be found in some other
applications such as air conditioning systems in airplanes and gas liquefaction plants but with
different objective, provide cold air. The project focuses on the studying of blading to match
a certain boundary conditions and developing the performance of the turbo-expander
components.

H1043 INDIVIDUAL PROJECT 1 CANDIDATE: 122511

 Table of Contents

STATEMENT OF ORIGINALITY ...........................................................................................................................3

ACKNOWLEDGMENT ........................................................................................................................................4
NOMENCLATURE .............................................................................................................................................5
CHAPTER 1: INTRODUCTION ............................................................................................................................6
CHAPTER 2: LITERATURE REVIEW.....................................................................................................................7
2.1 BACKGROUND PRINCIPLES ......................................................................................................................... 7
2.2 CENTRIFUGAL COMPRESSOR PERFORMANCE: ........................................................................................ 10
2.3 RADIAL INFLOW TURBINE PERFORMANCE:............................................................................................. 16
2.4 OPTIMUM DESIGN ..................................................................................................................................... 18
CHAPTER 3: METHODOLOGY..........................................................................................................................19
3.1 DESIGN GEOMETRY .................................................................................................................................. 20
3.2 PRELIMINARY TESTS................................................................................................................................. 21
3.2.1 Flow Rate Test ................................................................................................................................... 21
3.2.2 Frictional and Mechanical Losses: .................................................................................................. 23
3.3 BLADES PROFILING ................................................................................................................................... 26
3.4 EXPERIMENTAL SETUP.............................................................................................................................. 30
CHAPTER 4: RESULTS AND DISCUSSION .........................................................................................................31
CHAPTER 5: CONCLUSION ..............................................................................................................................35
5.1 RECOMMENDATIONS & FUTURE WORK ................................................................................................... 35
REFERENCES...................................................................................................................................................36
LIST OF FIGURES.............................................................................................................................................37
LIST OF EQUATIONS .......................................................................................................................................38
LIST OF TABLES ..............................................................................................................................................39
APPENDIX A .................................................................................................................................................... II
APPENDIX B ................................................................................................................................................... III
APPENDIX C ................................................................................................................................................... IV

H1043 INDIVIDUAL PROJECT 2 CANDIDATE: 122511

 Statement of Originality

I declare this dissertation has been written by me for my final year individual project. It does
not contain any materials that belong to other people. References have been included where
applicable.

H1043 INDIVIDUAL PROJECT 3 CANDIDATE: 122511

 Acknowledgment

First and foremost I would like to thank my supervisor Dr. Vasudevan Kanjirakkad for his
great support in this project and his supervision. Secondly, I would like to express my thanks
to the technician Simon Davies, Kevin Brady and the chief technician Peter Henderson for
their kind assistance and design recommendations. I also would like to thank Dr. Romeo
Glovnea for his recommendations. Last but not least, I would like to thank my family and all
my friends for their support and encouragement.

H1043 INDIVIDUAL PROJECT 4 CANDIDATE: 122511

 Nomenclature

Symbol Quantity Units

V Absolute Velocity m/s
β Blade Angle degree
α Flow Angle degree
ṁ Mass flow rate Kg/s
ω Rotational Speed – rad/sec
N Rotational Speed RPM
h Enthalpy, Blade Height J/Kg , mm
Cf Skin Friction Coefficient -
d Diameter mm
I Moment of Inertia of rotor Kg.m2
P Pressure Pascal
r Radius mm
T Torque N.m
W Shaft Power Watt
w Specific work Joule
ρ Density Kg/m3
μ Dynamic Viscosity m2/s
Ƞ Efficiency -
σ Slip Factor -
Z Number of blades -

Subscripts
is Isentropic
ov Overall
mech Mechanical
D Diffuser
r Radial
W Relative
⍬ Swirl
o Total or Stagnation
1 Compressor Impeller inlet
2 Compressor Impeller exit
3 Diffuser inlet
4 Diffuser exit
5 Nozzle inlet
6 Nozzle exit
7 Turbine rotor inlet
8 Turbine rotor exit

H1043 INDIVIDUAL PROJECT 5 CANDIDATE: 122511

 Chapter 1: Introduction

The objective of the project is to design and test a small turbo-expander that will be
manufactured from simple materials. The purpose of the project is to achieve the maximum
possible performance at a maximum rotational speed. The project will be focusing on blades
profiling of a centrifugal compressor and a radial inflow turbine to achieve requirement. In
order to do this, it is vital to introduce the components of turbo-expander such as the nozzle,
rotor disc, stator disc and axial shaft. The turbo-expander is driven by sucking air through the
compressor intake by a vacuum pressure system that is attached at the turbine exhaust (1).
The aim of the compressor is to increase the air pressure and temperature that is taken from
the atmosphere and delivers it to the turbine through transfer ports. The turbine will be used
to expand the compressed air and, therefore, some of its kinetic energy will be extracted to
drive the compressor and the rest can be used for other applications (2). When the rotational
speed is set to idle mode, the suction power from the vacuum pressure system will be
dissipated by frictional losses such as mechanical friction in the bearings and fluid friction in
the blades (3). This can be reduced by applying the knowledge of bearings selection.

The main objective is to obtain the highest possible ratio of the flow power at compressor
exit to the maximum suction power from the vacuum pressure system. This includes
designing the turbo-expander from dimensions to materials selection to meet boundary
conditions such as the rate of mass flow and pressure rise of the vacuum pressure system and
improving its efficiency.

H1043 INDIVIDUAL PROJECT 6 CANDIDATE: 122511

 Chapter 2: Literature Review

2.1 Background Principles

Turbo-expander Fundamental
The Term ‘Turbo-expander’ defines a single unit that consists of two primary components,
compressor and expander connected on a single shaft (2). The compressor is the driven unit
and the expander (turbine) is the power unit (2). Large Turbo-expanders are commonly used
for gas liquefaction and power generation (4). The efficiency of the turbo-machine depends
on its design, specifications and the user boundary conditions, for example, for high mass
flow rate the benefit will be in the axial flow compressors while the centrifugal compressors
used for a low mass flow rate, and likewise for the turbine (1). The centrifugal effects can
increase the specific work input to the fluid for a given dimensions and blades profile. The
centrifugal compressor stage consists of two primary components a rotating disc (impeller)
and a stationary disc (diffuser). Impellers are used to increase the energy of the working fluid,
the momentum of the fluid will be increased due to the centrifugal effect, whirling it radially
(5).The velocity of the fluid and its static pressure rise at this stage. Therefore, the kinetic
energy of the fluid will be transferred to pressure energy by the diffuser (5). The fluid will be
directed to the expander so that its temperature and pressure will decrease due to expansion.
Therefore, the kinetic energy of the fluid will be extracted to mechanical energy to drive the
shaft. Accordingly, the laws of thermodynamics and aerodynamics are involved in designing
and running turbo-expanders.

Centrifugal Compressor
This section of will be focusing on the centrifugal compressor and its stages. The essential
stages of the centrifugal compressor are:

 The eye of the impeller:

This is the inlet of the impeller just after the intake nozzle. The function of the eye is to
distribute the flow that sucked in through the system. The eye has an inducer or rotating
guide vanes, usually finishes when the flow turns from the axial direction to radial flow (1).
Some designers increase the inducer area to reduce the flow diffusion (1).

 The rotor disc (impeller):

This component is the heart of the centrifugal compressor as its increase the velocity of the
working fluid and rises its static temperature and pressure. Also, the energy level of the fluid
increases due to the increase of momentum in the fluid at the impeller (1). The impeller has a
series of curved vanes, shroud and hub, to guide the flow path. The shroud is the curved
surface that forms the outer boundary of the flow and the hub is the inner curved surface (1).
This means blades designing is an essential part for achieving high compressor efficiency.

H1043 INDIVIDUAL PROJECT 7 CANDIDATE: 122511

  The vaneless space:
Generally, centrifugal compressors are integrated with either vaned or vaneless diffusers to
convert the kinetic energy at the outlet of the impeller into static pressure.
The gap between the outlet of the impeller and the inlet of the diffuser called vaneless space
(1). Pressure rises in this space, hence, the larger the space the more pressure rise will occur
which might affect badly on the efficiency of the compressor (3).

 The stator disc (diffuser):

The diffuser contains a set of stationary blades that surround the impeller. They are designed
in a manner that the outlet flow at the impeller will encounter an increase in flow area as it
passes through them (5). As a result, this will cause a reduction in the flow velocity. The
purpose of the diffuser is to control the characteristics of the fluid by converting the kinetic
energy of the fluid into pressure energy, by diffusing the flow leaving the impeller, resulting
in an increase in the pressure energy (5). This conversion occurs due to Bernoulli’s principle.
𝑃1 + 1⁄2 𝜌 𝑉12 + 𝜌 𝑔 ℎ1 = 𝑃2 + 1⁄2 𝜌 𝑉22 + 𝜌 𝑔 ℎ2 (1)

Losses in stagnation pressure will occur if the diffusion rate is too high due to flow mixing
whereas in low diffusion rate the fluid will be exposed to a long wall resulting in fluid
friction losses. In general, the optimum rate of diffusion is 7o or 8o (1).

Figure 1.Centrifugal Compressor (1). Figure 2.CentrifugalCompressor stage (1).

(Used by Sir Frank Whittle) (With Kind Permission of Rolls-Royce plc)

H1043 INDIVIDUAL PROJECT 8 CANDIDATE: 122511

Radial Inflow Turbine
The radial flow turbine has similar components to the centrifugal compressor listed in the
sub-section 2.2. However, the flow path entering the turbine encounters different order of
stages than the compressor (4). I.e. the flow path enters the turbine through the stationary
vanes, resulting the flow to pass through the vaneless space and enters the turbine rotor. The
rotor has an inducer similar to the compressor. The purpose here is to receive an axial flow
for a certain distance at the inducer and converts it to radial flow at the end of it as the
shrouds are deflected at a set angle (5), resulting in expanding the fluid.
The turbine is the power unit as it expands the high pressure energy delivered from the
compressor (2). The expansion process occurs due to the shape of the blades, resulting in
extracting the kinetic energy from the fluid and converting it into mechanical energy to the
shaft (2). During the expansion process, the temperature of the fluid will cool down, the
pressure will decrease and the flow will accelerate. This means the boundary layers are more
stable and the blades can receive higher loading without boundary layer separation (1). The
turbine has an exhaust (outlet nozzle) for impulse or reaction purposes.

Figure 3. Radial inflow turbine

H1043 INDIVIDUAL PROJECT 9 CANDIDATE: 122511

2.2 Centrifugal Compressor Performance:
There are different methods to change the velocity of a working fluid by either changing the
area of the flow (Bernoulli’s) or stream curvature (blades). The flow in a centrifugal
compressor has three-dimensional motion analysis. These analysis are taken at inlet and
outlet cross-section area of the impeller and the same process can be used through the
machine (1). This section will illustrates velocity components and compressor performance.

Inlet Casing
The working fluid enters the machine with velocity Vo to V1 and with a static pressure drop
from Po to P1. In ideal conditions, the stagnation enthalpy is constant across the inlet (3).
1 1
ℎ𝑜 + 𝑉𝑜2 = ℎ1 + 𝑉12 ( 2)
2 2

The Impeller
There are different types of impellers such as one sided, two sided, shrouded, unshrouded,
radial, backward swept and forward swept vanes. Each type has different characteristics such
the arrangement of the blades, figure4 shows some samples of impeller types.

Figure 4.samples of impeller types

Flow enters the impeller with absolute velocity V, and has components of velocity, Vr, Va
and V⍬, radial, axial and tangential respectively as shown in Figure 4.
𝑉 2 = 𝑉𝑟2 + 𝑉𝑎2 + 𝑉⍬2 (3)

The angle made between the radial velocity and relative velocity is the exit blade angle β. If
we assumed that the impeller has no guide vanes and the flow enters the impeller axially, at
Va1, the velocity components Vr1 and V1 in this case will be equal to Va1. Thus, there will be
no swirl component and the angular momentum of the flow is zero (6). However, the flow
leaves the impeller with swirl velocity due to the centrifugal action, Vθ2 = U2 +V2 sinβ2 and
Vr2 = V2 cosβ2. In ideal conditions the tangential component at outlet is equal to the impeller
tip speed U2. Figure 5 shows the velocity triangle at inlet and outlet of the impeller.

H1043 INDIVIDUAL PROJECT 10 CANDIDATE: 122511

The static enthalpy change in centrifugal compressor is larger than the axial compressor, this
can be explained by the following equation:
1 1 1 2
ℎ2 − ℎ1 = (𝑉22 − 𝑉12 ) + (𝑈22 − 𝑈12 ) + (𝑉𝑊1 2 )
− 𝑉𝑊2 (4)
2 2 2
1
The first term in the right hand side of equation (4),2 (V22 − V12 ), represents the actual
1
increase in kinetic energy and the second term, (U22 − U12 ) is the contribution from the
2
2 1
centrifugal effect caused by the change in radius. The third term in the equation, 2 (VW1 −
2 ),
VW2 is the contribution from the diffusion of the relative velocity. Thus, in order to increase
the pressure in the compressor, the work transfer across the compressor must be large.
However, a high pressure ratio may lead to some aerodynamics issues such surge and stall.
Compressor efficiency and blade speed effect strongly on the pressure ratio (1).
𝑃𝑜3 (
𝛾
)
= [1 + (𝛾 − 1)Ƞ𝑐 𝜎(1 − 𝜙2 𝑡𝑎𝑛𝛽2′ )𝑀𝑢2 ] 𝛾−1 (5)
𝑃𝑜1

Where, 𝜙 = 𝑉𝑟 /𝑈 and Mu = U2/ao1 is the blade Mach number.

Because the velocity changes from the inlet to the outlet, the temperature and pressure of the
fluid will therefore change. In high pressure ratio compressors it might be beneficial to
introduce pre-rotation to the working fluid entering the impeller to reduce its relative
velocity.
Pre-whirl can be achieved at the inlet by fixing guide vanes to the casing, flow enters with an
angle α1, and this changes the velocity triangle at the inlet. Thus, angular deviation of the
absolute velocity can be realised at the leading edge of the impeller, Vθ1 . These guide vanes
act to change the pressure of the fluid at the first-stage before entering the impeller. Hence,
the air density decreases as the pressure drop increases. As a result, the compressor mass flow
production will decrease. Another advantage of the pre-whirl is to reduce the curvature of the
impeller vanes. Therefore, the inlet guide vanes need to be designed carefully to minimise
pressure losses and avoid some aerodynamics issues that will be discussed in the later stages.

𝑃 = 𝑇𝜔 = ṁ[𝑈2 𝑉⍬2 − 𝑈1 𝑉⍬1 ] (6)

From Euler’s Equation (6), Where U = ωr, the power of the compressor is related to the
tangential velocity at inlet. If the tangential velocity increases across a blade row, then the
angular momentum increases and it said the work is done on the fluid (7). This can be applied
for both rotating and stationary blades.

H1043 INDIVIDUAL PROJECT 11 CANDIDATE: 122511

The specific work done on the fluid will therefore be:
𝛥𝑊 = 𝑈2 𝑉𝜃2 − 𝑈1 𝑉𝜃1 (7)

Form the equation above, the work done on the fluid increases with the increase of the blade
speed U2 and Vθ2 (8). The second term of the equation can be neglected if the flow enters the
compressor without tangential component, α1 =0, else, the work done will decreases.
However, the flow will have swirl component at the outlet of the impeller, positive swirl
decreases the work and negative swirl increases the work (8).

Figure 5. Velocity triangle for radial compressor (9).

Slip Factor
In centrifugal compressors the relative flow at the exit of the impeller will does not receive
perfect guidance from the blades therefore it will deviates. The deviation in the exit angle is
called a slip. Hence, the slip factor can be introduced as
′
𝑉𝜃2 − 𝑉𝜃2
𝜎 =1− (8)
𝑈2
′
Vθ2 is the actual swirl velocity and Vθ2 is the ideal swirl velocity with no slip. The typical
range of slip factor suggested by (Seppo A. Korpela) is 0.83< σ <0.95 (8). The fall in the
swirl velocity at the impeller tip depends upon the number of blades on the impeller, ie. the
larger the number of blades the smaller will be the slip and Vθ2 approaches U2 (8) .
Vθ2
Thus,𝜎 = .
U2
According to Stanitz recommendation, the slip factor for radial compressors is defined as
0.63𝜋
𝜎 = 1− (9)
𝑍
Where Z is the number of blades. It is necessary to assume a slip factor at design point.
H1043 INDIVIDUAL PROJECT 12 CANDIDATE: 122511
The Diffuser
The diffuser can be either vaned or vanless space, or a combination of both. The main
purpose of the diffuser is to convert the exit kinetic energy into static pressure. Static pressure
and temperature increase with the decrease of the absolute velocity across the diffuser. When
the flow enters the diverging passage, diffuser, it reverse is direction and flow back to the
pressure gradient (5). Eddies phenomena may occurs if the divergence angle is too large, this
result in transferring the kinetic energy into internal energy and the useful pressure rise will
decrease (5). However, a small angle of divergence gives a high value of skin friction loss as
the diffuser blades will be long. An optimum angle of divergence of 7 o is suggested by
(Cohen&Roger) for long diffusers and twice this value can be used for low length/width
ratio.

In order to design a smooth flow process, the diffuser vanes angle at the leading edge must
suit the direction of the absolute velocity of the flow and the vanless space between the
impeller and the stator must be considered. The angular momentum at the diffuser is constant
as there is no energy is delivered to the flow after the impeller (5). Hence, Vθ2 from the
impeller decreases in the diffuser. Similarly, Vr2 from the impeller decreases in the diffuser as
the area of the flow is proportional to the radius in the radial direction, Continuity equation.
As a result, the absolute velocity of the flow will decreases and some diffusion occurs in the
vanless space (5).

Figure 6.Types of diffusers (9).

H1043 INDIVIDUAL PROJECT 13 CANDIDATE: 122511

The diffuser performance can be measured by the diffuser efficiency Ƞ D, the ratio of the
actual change in enthalpy to the isentropic change in enthalpy. Alternatively, the performance
can be measured by the total pressure recovery ratio Po2/Po1 (1).
ℎ2𝑠 − ℎ1 𝐶12 − 𝐶2𝑠
2
Ƞ𝐷 = (ℎ )
= 2 − ℎ1 = 2 ( 10)
ℎ2 − ℎ1 𝐶1 − 𝐶22

Also, for constant temperature process, 1 - 2

And for constant pressure process, 2s – 2
𝑇2𝑠 𝑇2
Ƞ𝐷 = ( )/( (11)
(𝑇1 − 1) (𝑇1 − 1)

𝑇2 𝑇2 𝑇2𝑠 𝑃𝑜1 𝑃2 (𝑦−1)/𝑦

= ( ) ( ) = [( ) ( )]
𝑇1 𝑇2𝑠 𝑇1 𝑃𝑜2 𝑃1

Substitute this in equation (12),

𝑃 𝑦−1/𝑦
(𝑃2 ) −1
1
Ƞ𝐷 = (12)
𝑃 𝑃 𝑦−1/𝑦
[(𝑃01 ) (𝑃2 )] −1
02 1

Figure 7.Mollier diagram for flow process in a compressor. (8)

H1043 INDIVIDUAL PROJECT 14 CANDIDATE: 122511

Mach Number
In high speed centrifugal compressors Mach number effects may become important with high
inlet velocity. The ratio of the velocity of a body to the speed of sound is called Mach
number.
𝑉1
𝑀= (13)
√𝛾𝑅𝑇1

Flow separation may occur in the suction surface of the impeller and break away from the
curved part of the vane (5). Hence, a shock wave might arise. It is possible to reduce the
relative velocity of the flow and Mach number by introducing pre-whirl at the impeller inlet.
However, this will reduce the work capacity of the compressor, equation (6).
High Mach numbers in the leading edge of the diffuser vanes imply high flow velocity and
large stagnation pressure which causes a variation in the static pressure (5). Hence, it will be
transferred in the radial direction through the vanless space. As a result, the purpose of the
vanless space is to avoid the shock losses and the static pressure variation.

In case of the velocity of the flow in the passage reaches the speed of sound, M=1, a large
flow blockage may occur due to the separation on the pressure surface or formation of shock
waves, therefore, the flow will choke at some cross-section (1). This means no further
increase in the mass flow either by increasing the rotational speed or by decreasing the back
pressure (1). Choke may occurs in the inlet, impeller and diffuser. Other phenomenon may be
considered when designing a turbomachine such as Surge and rotating stall.

Surging in compressor is related to the sudden drop in delivery pressure where the
compressor cannot add enough energy to overcome the system resistance or back pressure
(5). This causes a rapid reverse in the flow direction.

At a low flow rates, the reduction in radial velocity causes the flow to enter the impeller with
large positive incipience resulting in flow separation on the suction surface leading to rotating
stall (5). It also occurs at the diffuser due to large positive incidence at reduced flow rates.
Rotating stall may causes aerodynamic vibration resulting in fatigue to the machine parts.
Figure 8 shows how these phenomenon occur with respect to flow rate and pressure ratio at a
certain speed. Figure 9 explains the rotating stall on the suction surface.

H1043 INDIVIDUAL PROJECT 15 CANDIDATE: 122511

 Figure 8. Choking (5) Figure 9.rotating stall (5)

2.3 Radial Inflow Turbine Performance:

In radial turbines, the flow travels with a high tangential velocity and directed inwards to
leave the rotor with a small whirl velocity to minimise losses (5). The radial inflow turbine
looks similar to the centrifugal compressor but with a set of nozzle vanes instead of the
diffuser vanes and the velocity triangle in radial turbines are similar to those in the centrifugal
compressor. The flow enters the nozzle vanes with relative velocity equals to the radial
velocity, zero incidence, and leaves with relative velocity that has a tangential component
equals to the compressor𝑉𝜃2. The absolute velocity at the impeller exit is axial because 𝑉𝜃3is
zero.

Figure 10. Typical turbine vanes (8) Figure 11.velocity triangles at impeller (8)

H1043 INDIVIDUAL PROJECT 16 CANDIDATE: 122511

The specific work delivered by the turbine is,

1 2 1 1 2
𝑤= (𝑉2 − 𝑉32 ) + (𝑈22 − 𝑈32 ) + (𝑉𝑊3 2 )
− 𝑉𝑊2 (14)
2 2 2
From this equation, it can be noticed that the increase in the absolute velocity at the inlet can
increase the work. This can be obtained by orienting the nozzle vanes such that the flow
enters the rotor with a large angle α2 (8). Likewise, a small increase in the relative velocity
can effect positively on the work. This can be achieved by directing V W2 radially inwards at
the rotor inlet. The same reason leads V3 to be as small as possible and equals to Va3 at the
exit (10). In addition, the relative velocity at the exit increases by making the magnitude of β 3
1
large. The term 2 (𝑈22 − 𝑈32 ) in equation 14, can increase the work delivered by increasing U2
and decreasing U3.

𝑤 = 𝐶𝑝 (𝑇𝑜1 − 𝑇𝑜3 ) = 𝑉𝜃2 𝑈2 = 𝑈22 (15)

Equation 15 shows that 𝑉𝜃2 should be larger than 𝑉𝜃3, similarly, U2. The work done can be
increased by making 𝑉𝜃3negative. This may leads the flow to choke as β3 increases and the
exit relative Mach number becomes large (5).

If there is a wasted kinetic energy at the exit, total-to-static efficiency equation is used,
𝑇
1 − (𝑇𝑜3 )
𝑜1
Ƞ𝑡𝑠 = 𝛾−1 (16)
𝑃 𝛾
1 − (𝑃 3 )
𝑜1

Or
𝑇𝑜1 − 𝑇𝑜3
Ƞ𝑡𝑠 = (17)
𝑇𝑜1 − 𝑇3′

There are some useful parameter that we need to consider in designing a radial turbine (1),
𝑃𝑜1 −𝑃𝑜2
 Stagnation pressure loss coefficient: YN = ≈ ξN (1+0.5 γ 𝑀22 )
𝑃𝑜2 −𝑃2
ℎ2 −ℎ2𝑠
 Enthalpy loss coefficient : ξ N = 1 2
𝑉
2 2
𝑉
 Velocity coefficient : ϕN = 𝑉 2
2𝑠
These equations are applied at the nozzle, similarly, for the rotor but replacing V with VW (1).

H1043 INDIVIDUAL PROJECT 17 CANDIDATE: 122511

2.4 Optimum Design
This section will delivers the procedure of designing a turbomachinery by illustrating some
recommendations and limitations from the literature, eg Cohen & Roger (2001) and Whitield
& Baines (1990).
First of all, the parameter To1, Po1, ṁ and N are required in order to determine the design
dimension and then the correlated velocity diagrams through the machine.
𝑉𝑊2
According to de Haller ratio at the compressor impeller, > 0.75, to minimise the flow
𝑉𝑊1
separation in the impeller (11). The flow angle, α2, at the impeller exit effects on the stability
of the radial vanless diffuser, therefore, as a general rule, α2 > 70o. Thus, α2 is recommended
to be in the range of 60o to 65o at design point when the rotor diameter ratio is greater than
𝑈∗60
0.55 (12). The rotor diameter can be found by this relation, 𝑑 = . Blade width can be
𝜋∗𝑁
obtained by applying Continuity equation at the outlet (13). The number of blades can be
obtained by equating the slip factor ratio with Stanitz correlation, a typical blade number is
between 10 and 12 (14). Designing the diffuser need more care than the impeller as the flow
experienced some aerodynamics issues as mentioned in the earlier sections. The vanless
space must be considered in the calculation in order to get the correct flow angles and vanes
shape (1).

H1043 INDIVIDUAL PROJECT 18 CANDIDATE: 122511

 Chapter 3: Methodology

This chapter will focus on the design method of the centrifugal compressor and the radial
turbine. Precisely, we will be meeting blades design. For design purpose, it is necessary to
assume the efficiencies of the compressor and the turbine (3). Therefore, the pressure rise
across the compressor and the pressure drop across the turbine can be determined by using
the conventional definition of isentropic efficiency of the compressor and the turbine (3):
𝑚̇ ∆𝑃𝑐 ∆𝑃𝑡
= 𝑚̇𝜂𝑚𝑒𝑐ℎ 𝜂𝑡 (18)
𝜂𝑐 𝜌 𝜌
Where, ∆𝑃𝑐 = 𝑃𝑐𝑜𝑚𝑝 − 𝑃𝑎𝑡𝑚 and ∆𝑃𝑡 = 𝑃𝑐𝑜𝑚𝑝 − 𝑃𝑣𝑎𝑐

By rearranging equation (18),

∆𝑃𝑐 = 𝜂𝑚𝑒𝑐ℎ 𝜂𝑡 𝜂𝑐 ∆𝑃𝑡 = 𝜂𝑜𝑣 ∆𝑃𝑡 (19)
Where 𝜂𝑜𝑣 is the overall efficiency which is defined as:
∆𝑃𝑐
𝜂𝑜𝑣 = 𝜂𝑚𝑒𝑐ℎ 𝜂𝑡 𝜂𝑐 =
∆𝑃𝑡
Therefore, the objective is to maximise the function:
𝑚̇ ∆𝑃𝑣𝑎𝑐 (𝜂𝑜𝑣 /(1 − 𝜂𝑜𝑣 )
(20)
(𝑚̇ ∆𝑃𝑣𝑎𝑐 )𝑚𝑎𝑥

H1043 INDIVIDUAL PROJECT 19 CANDIDATE: 122511

3.1 Design Geometry
This section will illustrate the dimensions of the prototype model for this project.
Figure12 shows the cross section view of the
turbo-expander. The body is 440 mm2 and 186
mm deep. The shaft is 186 mm long and has a
diameter of 25 mm. Sealed bearings are
attached at both ends on the shaft. A Sleeve is
screwed on the foam body so that bearings are
securely locked from the inside and from the
outside by a washer. There are two discs have
been used in this design, a 200 mm outer
diameter rotor and a 310 mm outer diameter
stator with a 4 mm diameter vaneless space in
between. The stator disc is screwed to the foam
body for the sake of blades modification. Nuts
will be used for tightening the washer and the
rotor disc. Intake nozzle is attached on the left-
hand side in figure 12 and a turbine outlet is on
the other side.

Figure 12.cross section view

Figure 13 shows the transfer ports, 18x 30 mm. Also, it illustrates 8 holes with a 9 mm
diameter for attaching a cover plate. The cover plate is 10 mm thick and it is used to maintain
the flow.

Figure 13.Front view of the model Figure 14.Turbo-expander

H1043 INDIVIDUAL PROJECT 20 CANDIDATE: 122511

3.2 Preliminary Tests
This section will discuss the primary tests that carried out in order to determine the mass flow
rate for a maximum suction power from the vacuum cleaner available. Also, it is necessary to
estimate a value for the mechanical efficiency Ƞmech.

3.2.1 Flow Rate Test

The performance of the vacuum cleaner is tested by using it to suck air through a smooth pipe
to measure the flow rate, whilst varying the flow with a simple valve as shown in figure 15.

Figure 15.Flow Measuring Tube.

The objective is to measure the pressure drop across the inlet with a pipe that suits the
vacuum cleaner available. The pressure rise from downstream of the valve to atmosphere is
also measured. Hence, the flow rate can be obtained by assuming a discharge coefficient for
the intake close to unity and applying Bernoulli’s equation across it.
The design values for the turbo-expander, pressure rise and mass flow rate, are determined at
the maximum suction power.
Table 1 shows the test result. Where P1 is the static pressure at the inlet and P2 is the vacuum
cleaner pressure as shown in figure 15. Also, 0 turn means the valve is fully opened and 6
turns when its fully closed.
Valve P1 P2 Tatm Patm
Turn Pa Pa C mm.gH
0 -45 -7670 19.5 764.5
2 -43 -7850 19.5 764.5
4 -40 -8400 19.5 764.5
4.5 -38 -8630 19.5 764.5
5 -33 -9350 19.5 764.5
5.25 -28 -10040 19.5 764.5
5.5 -20 -11180 19.5 764.5
5.625 -13 -12050 19.5 764.5
5.75 -10 -12750 19.5 764.5
5.875 -5 -13500 19.5 764.5
6 -2.52 -14200 19.5 764.5
Table 1

H1043 INDIVIDUAL PROJECT 21 CANDIDATE: 122511

From the result in table 1, it can be noticed that the pressure values are negative due to the
vacuum cleaner suction, the pressure is below atmospheric. The pressure drop across the inlet
can be calculated as shown:
-45 + 2.52 = -42.48 Pa.

Also, from the atmospheric pressure and temperature, the density yields out as:
𝑎𝑡𝑚 𝑃
𝜌 = 𝑅×𝑇 , where Patm in bars, Tatm is in Kelvin and R is the specific gas constant ≈ 287
𝑎𝑡𝑚
J kg−1 K−1
∴ 𝜌 = 1.214 𝐾𝑔/𝑚3

Bernoulli’s equation is used across the inlet to obtain the mass flow rate,
𝑃1 𝑉12 𝑃2 𝑉22
+ = +
𝜌 2 𝜌 2
V1 = 0 and by rearranging the equation above,
2(𝑃𝑜 −𝑃)
V= √ = 8.364 m/s
𝜌

𝑄 = 𝑉 × 𝐴 = 0.01643 m3/s
Where A is the cross-sectional area of the intake is 0.001964 m2
Therefore, the mass flowrate can be found as, 𝑚̇ = 𝑄 × 𝜌 = 0.0199 Kg/sec

The pressure rise through the pipe is found to be 9317 Pa. However, another flow test was
done directly to the turbo-expander by connecting the vacuum cleaner to it without a tube in
between. Thus, Pvac = 617 Pa.
The vacuum cleaner performance is better presented in the graph shown below:
pressure rise (N/m^2) suction power (W)

16000 140

130
14000
120
Pressure rise (N/m^2)

110
Suction power (W)

12000
100

10000 90

80
8000
70

60
6000
50

4000 40
0.000 0.005 0.010 0.015 0.020 0.025
mass flow rate (Kg/s)

Figure 16.Vacuum cleaner characteristics

H1043 INDIVIDUAL PROJECT 22 CANDIDATE: 122511

The maximum point of suction power is the desired design point of the turbo-expander, P ≈
135 W. Comparing this with the vacuum cleaner manufacturer specifications, 1400 W, the
vacuum is only delivered 10% of its suction power, appendix B. This is may be due to the
restriction of the flow by using a flexible hose, bended, which connect the vacuum cleaner
with the flow measuring tube. Another reason, the tube outlet was not smooth contraction and
was 1.3 meter long which is longer than required. Furthermore, the vacuum cleaner
maximum suction switch does not work which means less suction power and less flow rate
are delivered by the vacuum cleaner.

3.2.2 Frictional and Mechanical Losses:

A run-down test has been carried out on the rotating assembly of the model, without the
blades, in order to determine the torque and the mechanical losses and therefore, the
mechanical efficiency. The mechanical losses are due to friction in the bearings and to the
windage losses on the rotating discs (11).
This can be done by spinning the discs up with an electrical drill screwed gently on the end of
the shaft. The datum rotational speed is measured by a tachometer that picking up a signal
from a reflective point on the disc, 1190 RPM.
Time Speed 𝛚 Figure 17 shows the disc run-down test result. It
sec rpm rad/s is difficult to obtain a perfectly smooth curve
without using higher speed with longer timing.
50 1190 124.6666667
40 869 91.03809524
30 480.5 50.33809524
20 237.2 24.84952381
10 111.2 11.64952381
Table 2

Run-down test
1400

1200 y = 27.894x - 259.24

R² = 0.9692
1000
Speed (rpm)

800

600

400

200

0
10 15 20 25 30 35 40 45 50
Time (sec)

Figure 17.Result from the disc run-down test

H1043 INDIVIDUAL PROJECT 23 CANDIDATE: 122511

The total frictional torque on each disc can be estimated by assuming a constant value of skin
friction coefficient Cf = 0.0025 (11), and using the equation provided,
𝑟2
𝑇 = ∫ (𝐶𝑓 𝜌𝜔2 𝑟 2 𝜋𝑟)𝑟𝑑𝑟 = 0.2𝜌𝜋𝜔2 𝐶𝑓 (𝑟25 − 𝑟15 ) (21)
𝑟1

Where r1 and r2 are the inner and outer radius of the disc.
Hence, T = 0.00029767 Nm per face. This torque is calculated from the datum speed, 1190
rpm. Windage and bearing torque can be obtained by:
𝑑𝜔
𝑇=𝐼 (22)
𝑑𝑡
Where I is the moment of inertia of the rotating assembly which is 5.597x10-5 Kg.m2,
neglecting the shaft inertia. The rotational speed is the design point speed but the maximum
datum speed that have been obtained from the test was 1190 rpm. Therefore this speed will
be used to calculate Twindage+bearing. The result will yield out as 0.00018822 Nm. Therefore, the
mechanical efficiency from the power loss at this speed will be approximately 90%.
𝑇×𝜔
Ploss = (1-( )) × 100
9.5488

Furthermore, the power lost by disc friction is inversely proportional to the rotational speed
cubed. This is shown by the graph below.

Power lost by disc friction

1

0.99995

0.9999
Pwer lost (W)

0.99985

0.9998

0.99975

0.9997

0.99965
0 500000 1000000 1500000 2000000
rotational speed (ω3) (rad/s)

Figure 18.Power lost by disc friction.

Hence, the power of both the compressor and the turbine are also proportional to ω3. As a
result, the power lost due to friction should be constant as the speed changes.

H1043 INDIVIDUAL PROJECT 24 CANDIDATE: 122511

Bearings are usually used to reduce mechanical friction. Their friction increases less rapidly
with speed and therefore, the net mechanical efficiency rises with speed. For machines
designing, it is important to select a suitable bearing type according to some parameters such
as load and speed. Sealed Ball bearings will be used as they do not need for self-lubrication
as well as to prevent damages from dust (15). The inner diameter of the bearing is 20 mm and
the outer diameter is 72 mm with a bore of 19 mm. Each bearing can resist a maximum
dynamic load rating of 30.8 KN and static load rating of 15 KN (15). The maximum
operation speed is 10,000 rpm for this type of bearings.

Aluminium sleeve
Shaft Washer

Body

Air gap

Air gap

Rotor

Bearing
Figure 19. Details of the bearing arrangement.

H1043 INDIVIDUAL PROJECT 25 CANDIDATE: 122511

3.3 Blades Profiling
The rotational speed is an important parameter to determine at the design point. This can be
done by applying Euler’s equation:
∆𝑃𝑜
∆ℎ𝑜 = = 𝜔 ∆(𝑟 𝑉𝜃 ) = 𝜔 (𝑟2 𝑉𝜃2 ) (23)
𝜌 𝜂𝑐
Where, 𝑉𝜃2 is the swirl velocity at impeller exit. It can be related to the slip factor in this
equation:
𝑉𝜃2 = 𝜎 𝜔 𝑟2 + 𝑉𝑟2 𝑡𝑎𝑛 𝛽2 (24)

From continuity equation, 𝑉𝑟2 Can be determined:

𝑚̇
𝑉𝑟 = (25)
(2 𝜋 𝑟 𝜌 ℎ)
Equation (23) can be applied at any radius. Thus, the only unknowns in equation (21) are 𝛽2
the impeller exit angle and the rotational speed𝜔. The rotational speed and the pressure rise
will increase in the impeller if 𝛽2 has a negative value, back sweep (3). Hence, the velocity
triangle at the exit of the compressor impeller can be used to determine the other velocity
components.
The inlet flow angle at the diffuser is similar to the outlet angel at impeller obtained
previously (3).

Another important parameter in the diffuser vanes is the throat area. The velocity at this area
can be determined by using the continuity equation (3). The pressure rise in the diffuser
before the throat must be chosen properly, a typical value is 30% (3). The vanless space can
be optimised by setting the diffuser leading edge 𝑟3 in the range 1.05 to 1.1× 𝑟2 (3). The
diffuser divergence angle should be chosen carefully, low angle, in order to achieve the
highest diffuser effectiveness (3). This angle can be obtained by the blades number and their
curvature.

The swirl velocity at the turbine rotor exit should be zero in order to minimise the leaving
loss and pressure loss at the exhaust (3). Hence, the inlet velocity at rotor can be determined
by the compressor torque i.e.

1
(𝑟𝑉𝜃 )𝑡𝑢𝑟𝑏𝑖𝑛𝑒 𝑟𝑜𝑡𝑜𝑟 𝑖𝑛𝑙𝑒𝑡 = (𝑟 𝑉𝜃 )𝐶𝑜𝑚𝑝𝑟𝑒𝑠𝑠𝑜𝑟 𝑟𝑜𝑡𝑜𝑟 𝑒𝑥𝑖𝑡 (26)
𝜂𝑚𝑒𝑐ℎ
It is beneficial to make the radius of the turbine rotor as large as the disc (3). This will reduce
the swirl velocity and increase the blades velocity at the leading edge resulting in less
forward swept (𝛽 positve). At the design point, the blade angle at the turbine rotor must allow
negative incidence.

H1043 INDIVIDUAL PROJECT 26 CANDIDATE: 122511

The acceleration of the flow at the turbine nozzle will be increasing as the throat area
decreasing. Equations (23) and (24) can be used to obtain the exit angle and the exit velocity
of the flow. Since the flow travels through the transfer ports with no swirl velocity, therefore,
the inlet flow angle in the nozzle should be radial (3).

The velocity triangles for the inlt and outlet of the compressor impeller are:
U2
Vθ2
U1
Vr2
VW1 Vr1 = Va1 V2
VW2
β1 β2 α2

Inlet Outlet

U1 = 𝜔r1 = 8.4 m/s U2 = 𝜔r2 = 21 m/s

Vr1 = 7.8 m/s Vr2 = 1.965 m/s
β1 = -47o assume, σ= 0.9
α1 = 0o Vθ2 = 1.71 m/s
β2 = -40o
α2 = -84o

The velcoity triangles for the inlet and outlet of the diffuser are:

α3 Vr4 = V4
V3
Vr3

Vθ3
Inlet outlet

A = 2π(Vanless gap+r3)h = 0.0078 m2 Vr4 = 1.89 m/s

Vr3 = 1.89 m/s β4 = 62o
Vθ3 = 1.644 m/s
Β3 = α2 = 84o

There is no swirl velocity at the diffuser outlet, hence, the blade exit angle is determined from
𝑟
the diffusion ratio. 𝑟3 > 75% to avoid separation.
4

H1043 INDIVIDUAL PROJECT 27 CANDIDATE: 122511

 The velocity triangls for the inlet and outlet of the turbine stator are

Vr5 = V5 Vr6
V6 α6

Vθ6

Inlet Outlet

Vr5 = Vr4 Vr6 = 1.89 m/s

Β5 = 0o Vθ6 = Vθ3
α5 = 0o β6 = 84o

The velocity triangls for the inlet and outlet of the turbine rotor are:

U7
Vθ7
Vr7
V8=Vr8
β8 VW8
VW7 β7 α7 V7

U8
Inlet Outlet

Vθ7 = Vθ2 Vθ8 = 0

β7 = 38o β8 = -62o
α8 =0

The exact blade shape can be obtained by applying linear interpolation theory at the inner and
outer radius of the disc. The blades shape is shown in figure 20 and 21.
Compressor Rotor Diffuser
ri Bi ri Bi
r1 0.040 -47 B1 r2 0.102 84 B2
rx2 0.045 -46.3636 By2 rx2 0.107 82 By2
rx3 0.051 -45.7273 By3 rx3 0.112 80 By3
rx4 0.056 -45.0909 By4 rx4 0.116 78 By4
rx5 0.062 -44.4545 By5 rx5 0.121 76 By5
rx6 0.067 -43.8182 By6 rx6 0.126 74 By6
rx7 0.073 -43.1818 By7 rx7 0.131 72 By7
rx8 0.078 -42.5455 By8 rx8 0.136 70 By8
rx9 0.084 -41.9091 By9 rx9 0.140 68 By9
rx10 0.089 -41.2727 By10 rx10 0.145 66 By10
rx11 0.095 -40.6364 By11 rx11 0.150 64 By11
r2 0.100 -40 B2 r3 0.155 62 B3

H1043 INDIVIDUAL PROJECT 28 CANDIDATE: 122511

 Turbine Rotor Stator
ri Bi ri Bi
r1 0.040 38 B1 r2 0.102 0 B2
rx2 0.045 28.90909 By2 rx2 0.107 7.636364 By2
rx3 0.051 19.81818 By3 rx3 0.112 15.27273 By3
rx4 0.056 10.72727 By4 rx4 0.116 22.90909 By4
rx5 0.062 1.636364 By5 rx5 0.121 30.54545 By5
rx6 0.067 -7.45455 By6 rx6 0.126 38.18182 By6
rx7 0.073 -16.5455 By7 rx7 0.131 45.81818 By7
rx8 0.078 -25.6364 By8 rx8 0.136 53.45455 By8
rx9 0.084 -34.7273 By9 rx9 0.140 61.09091 By9
rx10 0.089 -43.8182 By10 rx10 0.145 68.72727 By10
rx11 0.095 -52.9091 By11 rx11 0.150 76.36364 By11
r2 0.100 -62 B2 r3 0.155 84 B3

0.2
0.2

0.15
0.15

0.1
0.1

0.05 0.05

0 0

-0.05 -0.05

-0.1 -0.1

-0.15 -0.15

-0.2 -0.2
-0.2 -0.15
Figure 20.Turbine blades shape
-0.1 -0.05 0 0.05 0.1 0.15 -0.2 0.2 -0.15
Figure 21.Compressor blades shape
-0.1 -0.05 0 0.05 0.1 0.15 0.2

H1043 INDIVIDUAL PROJECT 29 CANDIDATE: 122511

3.4 Experimental Setup
The experiment was run by connecting the vacuum cleaner to the turbine exit, as shown in
figure 22. Pressures tapping are attached on certain points such as the intake nozzle and cover
plate, in order to measure the pressures and temperatures across the turbo-expander by
connecting them to a digital sensor which is linked to LABVIEW, appendix C.

Figure 22. Apparatus set-up

The blades are cut into strips to the desired length and glued on the discs by using Araldite.
Each blade was immersed in the glue and stuck on the discs, one following the other, they
have been left for approximately two to three hours to dry. A fillet of glue was added on the
blades to ensure blades stability.
The turbine rotor was painted with black colour and a reflective point is used in order to
measure the rpm using a tachometer. All components were assembled and pressure tappings
are connected to the digital sensor and ready for test.
The rotational speed was controlled by moving the vacuum cleaner for a certain distances
from the turbine outlet. This would not give the desired values for both the vacuum cleaner
suction power and the flow rate due to the fact that sucking air from the atmosphere
dissipates some of the suction power.

Figure 23.test rig

H1043 INDIVIDUAL PROJECT 30 CANDIDATE: 122511

 Chapter 4: Results and Discussion

In this chapter we will analyse and discuss the results obtained from the test. The data has
been recorded for different values of RPM, 130, 230, 353, 700, 1400 and 1850. We will be
analysing the highest rpm available in order to find the maximum suction power from the
vacuum cleaner, therefore, the maximum power flow at compressor exit.
The pressure tappings location are represented as shown in figure 24.

Figure 24.Pressure Tappings Arrangement

Speed P1 P2 P3 P4 P5 P6 P7 P8 P9
rpm Pa Pa Pa Pa Pa Pa Pa Pa Pa
130 -42.349 -123.58 -87.92 -289.42 -291.64 -293.15 -345.84 -407.24 -559.595

230 -56.062 -54.18 -105.51 -361.78 -364.585 -364.6 -432.92 -512.18 -708.81

353 -72.24 -196.72 -133.14 -451.6 -453.06 -455.74 -544.88 -652.09 -895.34

700 -118.339 -325.34 -170.82 -646.21 -649.64 -655.29 -798.66 -1004.37 -1349.23

1400 -208.023 -585.1 -211.01 -973.05 -978.94 -987.86 -1246.5 -1673.95 -2248.3

1850 -289.927 -824.69 -200.96 -1195.11 -1207.24 -1221.89 -1587.36 -2253.58 -2975.84
Table 3.Pressure measurements

H1043 INDIVIDUAL PROJECT 31 CANDIDATE: 122511

P1 is the static pressure at the compressor inlet
P2 is the pressure at the rotor inlet
P3 is the pressure at the rotor outlet/diffuser inlet
P4 is the pressure at the diffuser outlet
P5 is the pressure at the transfer hole
P7 is the pressure at the stator outlet/rotor inlet
P8 is the pressure at the rotor exit
P9 is the pressure at the turbine outlet

From table 3, it can be noticed that the pressure drop across inlet equals to -247.578 Pa. This
means the flow rate is greater than the design point value, -42.48 Pa. Thus, the mass flow rate
from this experiment is 39.5% higher than the design point value, 0.0482 Kg/s. However, this
value can be increased if a proper tube with integrate valve is used to control the flow rate.

The pressure change across the compressor is given by:

∆𝑃𝑐 = 𝑃4 − 𝑃2 which gives a pressure change of -370.42 Pa.
And the pressure change across the turbine is also given by:
∆𝑃𝑡 = 𝑃8 − 𝑃6 which gives a pressure change of -1031.69 Pa
Thus, the overall efficiency is equal to 35.9%

𝑚̇ ∆𝑃𝑣𝑎𝑐 (𝜂𝑜𝑣 /(1−𝜂𝑜𝑣 )

The objective is to maximise , where ∆𝑃𝑣𝑎𝑐 = 𝑃𝑎𝑡𝑚 − 𝑃𝑣𝑎𝑐
(𝑚̇ ∆𝑃𝑣𝑎𝑐 )𝑚𝑎𝑥
By substituting values in, ∆𝑃𝑣𝑎𝑐 = 104900.84 Pa.
And the ratio of the flow power at compressor exit to the maximum suction power from the
vacuum cleaner yields as 56%, however, the flow has a drop in pressure at P3, telling that the
diffuser is not performing as required, this is can be shown in figure 25. The role of the
diffuser has been explained in chapter 2.

Flow Process In The Turbo-expander

1400
1195.11 1207.24 1221.89
1200
Pressure Changes (Pa)

1000
824.69
800

600

400 289.9265

200 200.96

0
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Rotational Speed (rpm)

Figure 25.Pressure Changes

H1043 INDIVIDUAL PROJECT 32 CANDIDATE: 122511

If the change in pressure at the rotor of the compressor is given by, P3-P2 = +623.73 Pa and
the change in pressure in the diffuser equals to -994.15 Pa. This proves that the diffuser is not
increasing the static pressure, hence, only the rotor pressure is transferring to the turbine. If
we assume that the diffuser is performing, therefore, Protor = -623.73 Pa, meaning that the
rotor increases the pressure to -370.42 Pa and the diffuser increases this to another
approximately 300 Pa.
Similarly, this can be shown by plotting the dimensionless ratio of the change in pressure
across the compressor and the change in pressure at the vacuum cleaner:

0.004

0.0035
(∆𝑃_𝑐𝑜𝑚𝑝)/(∆𝑃_𝑣𝑎𝑐 )

0.003

0.0025

0.002

0.0015

0.001
0 200 400 600 800 1000 1200 1400 1600 1800 2000
RPM

∆𝑃𝑐𝑜𝑚𝑝
Figure 26. The ratio of ∆𝑃𝑣𝑎𝑐

Furthermore, the maximum suction power from the vacuum cleaner is directly proportional to
the rotational speed as shown in figure 27.
5100

5080
maximum suction power

5060

5040

5020

5000

4980

4960
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Rotational speed (rpm)

Figure 27. Proportional relationship between the maximum suction powers with the RPM.

H1043 INDIVIDUAL PROJECT 33 CANDIDATE: 122511

The turbine exit blades are designed to eliminate the swirl but this was not achieved during
the experiment. Figure 29 shows that the flow is swirling at the exit by dripping Cole-Parmer
in the outlet pressure tapping. This could be due to the inaccuracy in the blades angle.

Figure 28. Swirl in the flow

In comparison with the experiment that carried out in Whittle Laboratory, our design values
are much lower than their values. Their mass flow rate is 0.1 Kg/s with a vacuum suction of
8000 Pa and a rotational speed of 8000 rpm, however, their blades fail at around 6000 rpm.
Our design rotational speed is 2000 rpm and we run it up to 1800 rpm.

H1043 INDIVIDUAL PROJECT 34 CANDIDATE: 122511

 Chapter 5: Conclusion

The project objectives have been met, however, the design point values for the flow rate and
pressure rise are lower than the expected, 0.1 Kg/s. This is due to the restriction in flow. The
turbine blades are designed to eliminate the whirl at the exit, however, the experiment proves
the opposite. This is could be possibly caused by the inaccuracy in the inlet and outlet blade
angles. The ratio of the flow power at compressor exit to the maximum suction power from
the vacuum cleaner is found as 56%, with an overall efficiency of 35.9%. The power lost by
disc friction is found to inversely proportional to the rotational speed cubed. In comparison
with Whittle Laboratory test, our mechanical efficiency of 90% is higher than what they
obtained, 85% (3). The test was run continuously for approximately 40 minutes at different
rotational speeds.

5.1 Recommendations & Future work

In my point of view, it would be beneficial to visualise the flow at the inlet and the outlet of
the turbo-expander in order to study the flow behaviour. Also, if time permits a different
blades shape can be test therefore, different result will yield out. It is better to measure the
flow rate with a proper measuring tube in order to get a better result. It would be ideal if the
blades were manufactured as the glue cannot resist a high centrifugal load, greater than 100
N. I also recommend to run the test with a more power full vacuum cleaner in order to obtain
a higher rotational speed with higher suction power.

H1043 INDIVIDUAL PROJECT 35 CANDIDATE: 122511

 References

1. Hall, S.L. Dixon and C.A. Fluid Mechanics and Thermodynamics of Turbomachinery.
s.l. : Elsevier, 2010. 978-1-85617-793-1.
2. FUNDAMENTALS OF TURBOEXPANDERS. SIMMS, JAMES. CALIFORNIA :
SIMMS MACHINERY INTERNATIONAL, INC, 2009.
3. The Turboexpander. Denton, J D. Cambridge : The American Society of Mechanical
Engineering, 1996.
4. Earl Logan, Jr. Handbook of Turbomachinery. New York : MARCEL DEKKER, INC. ,
2003. 0-8247-0995-0.
5. Saravanamuttoo, H. I. H. Gas Turbine Theory. s.l. : Prentice Hall, 2008. 0132224372.
6. M.White, Frank. Fluid Mechanics. New York : McGraw-Hill, 2008. 978-0-07-352934-9.
7. Kundu, Pijush K. Fluid mechanics. s.l. : ELSEVIER, 2012. 978-0-12-382100-3.
8. Korpela, Seppo A. Principles of Turbomachinery. Canada : John Wiley & Sons, 2011.
978-0-470-53672-8.
9. Khan, Rama S. R. Gorla & Aijaz A. Turbomachinery Design and Theory. s.l. : Marcel
Dekker, 2003. 0-8247-0980-2.
10. TURTON, R.K. Principles of Turbomachinery. England : CHAPMAN & HALL, 1995.
0 412 60210 5.
11. Glassman, A. Turbo design and Application. s.l. : NASA, 1975. Vol. 1.
12. Peng, William W. Fundamentals of Turbomachinery. s.l. : John Wiley & Sons, 2008.
13. Wilson D.G., Korakianitis Th. The design of high efficiency turbomachinery and gas
turbines. 1998.
14. Cumpsty, N. A. Compressor aerodynamics. s.l. : Longman Scientific & Technical, 1989.
9780582013643.
15. Childs, Peter R. N. Mechanical Design. s.l. : Butterworth-Heinemann, 2003.
0750657715.
16. Whitfield, A & Baines,. Design of Radial Turbomachines. N.C : Longmans, 1990.

H1043 INDIVIDUAL PROJECT 36 CANDIDATE: 122511

 List of Figures

Figure 1.Centrifugal Compressor (1). Figure 2.CentrifugalCompressor stage (1). .........8

Figure 3. Radial inflow turbine .................................................................................................9
Figure 4.samples of impeller types .........................................................................................10
Figure 5. Velocity triangle for radial compressor (9). ............................................................12
Figure 6.Types of diffusers (9). ..............................................................................................13
Figure 7.Mollier diagram for flow process in a compressor. (8) ............................................14
Figure 8. Choking (5) Figure 9.rotating stall (5) ...........................................................16
Figure 10. Typical turbine vanes (8) Figure 11.velocity triangles at impeller (8) ..16
Figure 12.cross section view ...................................................................................................20
Figure 13.Front view of the model Figure 14.Turbo-expander ........................................20
Figure 15.Flow Measuring Tube. ............................................................................................21
Figure 16.Vacuum cleaner characteristics ..............................................................................22
Figure 17.Result from the disc run-down test .........................................................................23
Figure 18.Power lost by disc friction. .....................................................................................24
Figure 19. Details of the bearing arrangement. .......................................................................25
Figure 20.Turbine blades shape Figure 21.Compressor blades shape ............................29
Figure 22. Apparatus set-up ....................................................................................................30
Figure 23.test rig .....................................................................................................................30
Figure 24.Pressure Tappings Arrangement .............................................................................31
Figure 25.Pressure Changes ....................................................................................................32
Figure 26. Whirl component in the flow .................................................................................34
Figure 27. Turbine.................................................................................................................... iv
Figure 29.Digital sensor ........................................................................................................... iv
Figure 28. Compressor .............................................................................................................. iv

H1043 INDIVIDUAL PROJECT 37 CANDIDATE: 122511

 List of Equations

𝑃1 + 12𝜌 𝑉12 + 𝜌 𝑔 ℎ1 = 𝑃2 + 12𝜌 𝑉22 + 𝜌 𝑔 ℎ2 (1) .............................................. 8

ℎ𝑜 + 12𝑉𝑜2 = ℎ1 + 12𝑉12 ( 2) ....................................................................................... 10
𝑉2 = 𝑉𝑟2 + 𝑉𝑎2 + 𝑉⍬2 (3) ............................................................................................... 10
ℎ2 − ℎ1 = 12𝑉22 − 𝑉12 + 12𝑈22 − 𝑈12 + 12𝑉𝑊12 − 𝑉𝑊22 (4) .............................. 11
𝑃𝑜3𝑃𝑜1 = 1 + 𝛾 − 1Ƞ𝑐𝜎1 − 𝜙2𝑡𝑎𝑛𝛽2′𝑀𝑢2(𝛾𝛾 − 1) (5)............................................... 11
𝑃 = 𝑇𝜔 = ṁ𝑈2𝑉⍬2 − 𝑈1𝑉⍬1 (6) .................................................................................. 11
𝛥𝑊 = 𝑈2𝑉𝜃2 − 𝑈1𝑉𝜃1 (7) ................................................................................................. 12
𝜎 = 1 − 𝑉𝜃2′ − 𝑉𝜃2𝑈2 (8) .............................................................................................. 12
𝜎 = 1 − 0.63𝜋𝑍 (9) ............................................................................................................ 12
Ƞ𝐷 = ℎ2𝑠 − ℎ1ℎ2 − ℎ1 = ℎ2 − ℎ1 = 𝐶12 − 𝐶2𝑠2𝐶12 − 𝐶22 ( 10)......................... 14
Ƞ𝐷 = (𝑇2𝑠(𝑇1 − 1))/(𝑇2(𝑇1 − 1) (11) ......................................................................... 14
Ƞ𝐷 = 𝑃2𝑃1𝑦 − 1/𝑦 − 1𝑃01𝑃02𝑃2𝑃1𝑦 − 1/𝑦 − 1 (12) .......................................... 14
𝑀 = 𝑉1𝛾𝑅𝑇1 (13) ............................................................................................................. 15
𝑤 = 12𝑉22 − 𝑉32 + 12𝑈22 − 𝑈32 + 12𝑉𝑊32 − 𝑉𝑊22 (14) ...................................... 17
𝑤 = 𝐶𝑝𝑇𝑜1 − 𝑇𝑜3 = 𝑉𝜃2𝑈2 = 𝑈22 (15) ...................................................................... 17
Ƞ𝑡𝑠 = 1 − 𝑇𝑜3𝑇𝑜11 − 𝑃3𝑃𝑜1𝛾 − 1𝛾 (16) ........................................................................ 17
Or Ƞ𝑡𝑠 = 𝑇𝑜1 − 𝑇𝑜3𝑇𝑜1 − 𝑇3′ (17) ..................................................................... 17
𝑚𝜂𝑐 ∆𝑃𝑐𝜌 = 𝑚𝜂𝑚𝑒𝑐ℎ𝜂𝑡 ∆𝑃𝑡𝜌 (18) .................................................................................... 19
∆𝑃𝑐 = 𝜂𝑚𝑒𝑐ℎ 𝜂𝑡 𝜂𝑐 ∆𝑃𝑡 = 𝜂𝑜𝑣 ∆𝑃𝑡 (19).......................................................................... 19
𝑚 ∆𝑃𝑣𝑎𝑐 (𝜂𝑜𝑣 /(1 − 𝜂𝑜𝑣) (𝑚 ∆𝑃𝑣𝑎𝑐)𝑚𝑎𝑥 (20) .............................................................. 19
𝑇 = 𝑟1𝑟2𝐶𝑓𝜌𝜔2𝑟2𝜋𝑟𝑟𝑑𝑟 = 0.2𝜌𝜋𝜔2𝐶𝑓𝑟25 − 𝑟15 (21) ................................................ 24
𝑇 = 𝐼𝑑𝜔𝑑𝑡 (22) ..................................................................................................................... 24
∆ℎ𝑜 = ∆𝑃𝑜𝜌 𝜂𝑐 = 𝜔 ∆𝑟 𝑉𝜃 = 𝜔 𝑟2 𝑉𝜃2 (23) .................................................................. 26
𝑉𝜃2 = 𝜎 𝜔 𝑟2 + 𝑉𝑟2𝑡𝑎𝑛𝛽2 (24) ........................................................................................ 26
𝑉𝑟 = 𝑚(2 𝜋 𝑟 𝜌 ℎ) (25) ...................................................................................................... 26
𝑟𝑉𝜃𝑡𝑢𝑟𝑏𝑖𝑛𝑒 𝑟𝑜𝑡𝑜𝑟 𝑖𝑛𝑙𝑒𝑡 = 1𝜂𝑚𝑒𝑐ℎ𝑟 𝑉𝜃𝐶𝑜𝑚𝑝𝑟𝑒𝑠𝑠𝑜𝑟 𝑟𝑜𝑡𝑜𝑟 𝑒𝑥𝑖𝑡 (26) ............................ 26

H1043 INDIVIDUAL PROJECT 38 CANDIDATE: 122511

 List of Tables

Table 1 ..................................................................................................................................... 21
Table 2 ..................................................................................................................................... 23
Table 3 ...................................................................................................................................... ii
Table 4 ...................................................................................................................................... ii

H1043 INDIVIDUAL PROJECT 39 CANDIDATE: 122511

 Appendix A

Project Management
First of all, we are going to learn the fundamentals of turbo expander and the main concept of
it, the physics and theories. After that, we will start designing the model on paper and decide
what components are involved. Next, we will choose the right dimensions for each part of the
turbo-expander such as the shaft, stator and rotor. Also, it is important that we take clearance
between every part into account so that we avoid as much friction losses and design failure as
we can. After that, we will start building the whole thing on CAD.
In term 2, we will assemble all the components together. After that, we will be making the
blades profile so that the turbo expander will be tested and the results will be analysed.
Improvement will be done if necessary for the design for a better performance. Preparation
for the presentation will be done as well as posters for the exhibition will be published.
Finally, the writing will take over. Weekly meeting with the supervisor is required for
tracking progress and improvement. This meeting is set to be every Tuesday.
Project requirements:
Equipment Cost (£)
#8x 25mm Self-tapping screw (180-695) 9.69
#8x 45mm Self-tapping screw (797-6165) 9.37
M10 Nuts (527-628) 8.19
Ball bearings (6404) 30 each
PLA 20
Foam 10
Cover Plate 44.62 each
Table 4

Components to manufacture Quantity

Shaft 1
Stator Disc 2
Rotor Disc 2
Aluminium Sleeve 1
Washer 2
Body 1
Bearing 2
Aluminium sheet 12 (1 meter each)
Table 5

H1043 INDIVIDUAL PROJECT ii CANDIDATE: 122511

 Appendix B

This is the vacuum cleaner available and used in performing

this project. Here are some manufacturer specifications

H1043 INDIVIDUAL PROJECT iii CANDIDATE: 122511

 Appendix C

Figure 29. Turbine

Figure 30. Compressor

Figure 31.Digital sensor

H1043 INDIVIDUAL PROJECT iv CANDIDATE: 122511

View publication stats